[E-voting] Irish Times today - Friday 3/12/2004 - about preferendums and e-voting

Peter Emerson pemerson at deborda.freeserve.co.uk
Mon Mar 22 16:16:55 GMT 2004

Dear Catherine,

Do I take it my letter was published in the IT on Friday 12/3/2004?  As I
mentioned in my other e-mail, I was in Russia, so I'm a little adrift...

Meanwhile, may I ask the said John Lambe to look at our de Borda web-site.
I've been accused of being a fenian bastard and quite a few other things
over the years, but nobody else has ever suggested I don't understand

Best wishes,


Peter Emerson
Director, The de Borda Institute
36 Ballysillan Road
Belfast BT14 7QQ
N Ireland
Tel +44 (0)28 9071 1795
pemerson at deborda.org

----- Original Message -----
From: "Catherine Ansbro" <cansbro at eircom.net>
To: <e-voting at lists.stdlib.net>
Cc: "Peter Emerson" <pemerson at deborda.org>
Sent: 12 March 2004 13:37
Subject: Re: [E-voting] Irish Times today - Friday 3/12/2004 - about
preferendums and e-voting

Ahhh.  You didn't realize what the writer meant.  He's not really talking
about e-voting per se, rather about a potential method for using it to make
*preferendum-style* decision making easier.  I agree with his point, AND I
can understand why someone not familiar with preferendum-style
decision-making would not understand his letter.  I'll do my best to
explain, because it's an area I'm quite interested in and because it would
be a great use of a good e-voting system having VVAT.  (Also it raises
really interesting possibilities for a GOOD poll on e-voting.)

Just to clarify what Peter Emerson is talking about--in a *preferendum*,
you'd have at least 3 options to choose from, and you would rank your
choices.  (Initially this seems similar to PR-STV.)  Votes are tallied in a
way that encourages people to rank all the available options--see below
(the *tally method* differs considerably from PR-STV).  It's crucial that
the list of options has been created in a democratic way, to include the
group's possible range of preferences.

Consider a preferendum having the following 8 options, in which the voter
is instructed to rank all 8 options (using 1 for first choice, 2 for
second, etc., using each number only once):

A - I prefer voting in a polling station using only a paper ballot on which
I write the numbers.
B - I prefer voting in a polling station using a computer or machine only
to print my numbered preferences on a paper ballot (so that my writing can
be accurately read by humans).
C - I prefer voting in a polling station using a paper ballot that is
counted by a scanning device.
D - I prefer voting in a polling station using a computer to record the
vote and count it, without using any paper at all.
E - I prefer voting in a polling station using a computer to record the
vote and count it, with a voter-verified paper trail (e.g., voter has seen
the printed paper ballot before casting their vote).
F - I prefer to vote over the internet, without a voter-verified paper
G - I prefer to vote by postal vote using a paper ballot on which I write
the numbers.
H - I prefer to vote in a polling station using a computer (whether or not
there is a voter-verified paper trail.)

(The list of options is incredibly important, and should be determined in a
democratic way.  I've put this together as a suggestion only, based on my
understanding of the options & issues we've discussed, trying to use
neutral wording.  There are problems with this list e.g., B & C are not
mutually exclusive, but it'll give you an idea of how it might be done.)

I would fill out the form by putting a number from 1 to 8 beside each of
these options, "1" for my first choice, etc.  (So for example I might put
"1" to G, "2" to B, "3" to C, etc. "6" for option H, "7" by option D and
"8" by option F   On the other hand, the current government policy might
rank D as "1", H as "2", not sure about the rest.)

Votes are tallied differently than in PR-STV.  Rather than having votes
transfer, votes are allotted "points" according to how the voter has ranked
them.  When the votes are tallied, if I have ranked all 8 options then my
first choice will get 8 points, my second choice will get 7 points,
etc.  If I only ranked 5 of the 8 choices then my first choice will get 5
points, my second choice will get 4 points, etc.  If I was trying to
"throw" the vote and only put down my first preference, that option will
get 1 point.  (You can probably start to see how this differs from PRSTV.)

When the votes are tallied you find out where the group consensus is.  It
is a very practical way of choosing from a range of options.

Peter Emerson, who wrote the Irish Times letter, is involved with the De
Borda Institute which has a number of excellent publications that  explain
various voting methods and how they affect the results.  There are lots of
examples in his books.  He has used the DeBorda method with groups all over
the world where there is strong polarization of views, e.g. explosive
political divisions as we see in Northern Ireland, the Balkans, etc.

We sometimes use this style of voting within the Green Party (e.g., for
choosing our Party Leader, or for local groups to indicate what issues
they'd like to prioritize in the coming European or local elections,
etc.)  It is a very democratic way of decision-making, when it is done

This kind of preferendum method would be greatly assisted by
computerization, because it makes the tallying process much easier.  (As
with PR-STV, I'd want to see that any computerization includes VVAT so that
the voter knows that his/her vote was recorded accurately and to facilitate
recounts if there are any questions raised.)  I know that Peter Emerson has
a simple computer program for tallying different kinds of counting methods,
but I don't know how much thought he has given to the importance of
VVAT.  Perhaps because he's seen to be a neutral outsider with integrity,
the potential for tampering with the counting software may not have
arisen.  But I'm sure he'd get the point immediately if someone suggested
that one side in a conflict should be entrusted with designing the voting
or counting software.  I don't know if he's considered the implications of
hardware failures that can cause a bit to flip, changing the tallies in
ways that may not necessarily be obvious.

To go back to the beginning, having a good list of options that accurately
reflects voters' possible preferences is very important.  That is also why
our Irish referenda are such a mess--we only have a Yes or No option.  So
in the constitutional amendment referenda on Divorce, Abortion, Nice
treaty, etc., we can't really tell how the electorate felt.  Did they vote
No because the proposed amendment was too lenient?  too
restrictive?  Because there were certain parts they objected to?  If our
constitution permitted a referendum in which we could choose from a range
of options we'd have a much more democratic result.  A DeBorda style count
is one way of doing this very effectively.

Coming back at last to ICTE, if we had an on-line poll and we used the list
of options A-H above (or some improvement of this list, e.g., I left out
any information relating to cost), and a DeBorda style vote tally, we would
get a lot more information about our group preferences.  (It would make
life much easier for political scientists like Karen, or for politicians
who want to know what people really think.  As I see it, part of the
problem in politics is that often there is no desire to know what people
really think, rather it is an exercise in manipulation to achieve a certain
pre-determined result.)  It would be very interesting to see the results of
such a poll among ICTE members, for example.  It would have been
interesting if all members of the Joint Oireachtas Committee had done a
preferendum using the above list of options, rather than a simple yes/no
vote on the currently proposed system.


   E.g., At 04:02 12/03/2004 -0800, John Lambe wrote:
>There isn't anything very important.
>There was one letter on electronic voting.
>It's not really relevant and somewhat nonsensical (for eaxmple, the
>author seems to think that you cannot have multiple preferences with a
>non-electronic voting system and doesn't seem to understand PR-STV).
>The author is in favour of some sort of audit.
>"Madam, - Amid all the noise about electronic voting, most people are
>talking about the voters, and few are discussing the elected
>representatives. Yet if the voters are to cast their preferences
>electronically in elections, could not the TDs do the same in
>An experiment on electronic voting was held in Belfast in 1991.
>...the 100 participants voted. A computer then displayed both the
>voters' audit and the result: the option with the highest average
>This, surely, is the great advantage of electronic voting: that it can
>cope with multi-option preference voting and thus cater for a plural
>There must, of course, be a paper (or electronic) audit, and everyone
>must be aware of the counting procedures involved.
>In decision-making, as in any rankings system, we identify the option
>with the highest average preference. In elections, because of the
>complications of PR-STV, the count should be done in stages, and all
>transfers should be done on the basis of a full (rather than a random)
>But first things first. Let's start by modernising the Dáil (and the
>EU). - Yours, etc.,
>PETER EMERSON, Director, The de Borda Institute, Ballysillan Road,
>Belfast 14."
>(Doesn't the Dail already have an electronic voting system).
>There were comments on electronc voting in other articles:
>Finance section: Article "Women need to get noisy on technology issues"
>'... politicians, as we've seen with the rather stilted reasons the
>Government has offered for pursuing electronic voting - often leave
>such debates to "the experts".'
>'The e-voting debate, such as it is, is not unconnected to what has
>happened with broadband. I listened to the Minister responsible, Mr
>Cullen, defend our rush towards e-voting with a comment to the effect
>of: "But maybe for once we can be first in the world to do something
>like this."
>Let's set aside the fact that this is hardly an adequate reason for
>doing anything, especially something as controversial as switching an
>entire voting process to a computer-based system.'
>'But get involved, prompt discussion and debate, and look at what has
>happened with e-voting - this rather dry topic has suddenly engaged the
>nation, and because of citizen disquiet, the State has at least been
>forced into a rethink, making some initial concessions.'
>Article: "Fears voiced over 'race card' in elections"
>Referring to Terry Leyden (FF) speaking about holding the referendum on
>citizenship rights on the same day as the elections:
>"He said further confusion would be caused in terms of electronic
>John Lambe
>Phone (mobile): +353 86 2895286
>Phone:          +353 1 4905842
>Address:        64 Brighton Road, Rathgar, Dublin 6, Ireland
>Email:          jlambe at johnlambe.com
>E-voting mailing list
>E-voting at lists.stdlib.net

More information about the E-voting mailing list